* Clear
  clear  
  
* Working Directory
  cd "~/Replication Materials for The Domestic Economic Costs of Sanctions - A Firm Level Analysis/"

* Log File
  log using "thedomesticcostsofsanctions.log", replace   
  
* Create Directory for Latex Tables
  mkdir rawtables
  
  * Note: The files in "/rawtables" are used to create the tables that appear in the
  * appendix but the tables have to be edited. The portmanteau tests were added
  * personally and the tables were edited fore aesthetics.
  
* Load Data
  use "rcs.dta"
  
* Set Data
  tsset date  
  
* Create Counter  
*  gen t = _n (Note: This is the code used to create the variable t in the data set. 
  
* Appendix Tables (Section 4)

  * Eli Lilly (1989)
  
    * Generate Returns
	  gen lly_returns = ln(lly_close/lly_close[_n-1])
    
	* Sanctions
	  gen tsanction3 = 1 if date > td("04jun1989") & date < td("25may1990")
	  recode tsanction3(.=0)
	
    * Bond Market Crash
	  gen f13minicrash = 1 if date > td("11oct1989") & date < td("17oct1989")
	  recode f13minicrash(.=0)
	  
    * Limit Time Period
      keep if date > td("30dec1988") & date < td("01jun1990")
		
    * Set for analysis
	  tsset t
	  
	* Table A.1 Models
	  eststo clear
	  
	  eststo: arch lly_returns, ar(1)
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch lly_returns, ar(4)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch lly_returns, ar(4) arch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch lly_returns, ar(4) arch(1) garch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch lly_returns, ar(4) arch(22)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch lly_returns, ar(4) arch(1) het(tsanction3 f13minicrash)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch lly_returns, ar(4) arch(1,22) het(tsanction3 f13minicrash)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch lly_returns, ar(4) arch(16,22) het(tsanction3 f13minicrash)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch lly_returns, ar(4) arch(22) het(tsanction3 f13minicrash)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Eli Lilly 1989) nodep
	  esttab using rawtables/lilly89.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Eli Lilly 1989) nodep replace
      
  * Pfizer (1989)
    
	* Clear
	  clear
	  
	* Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen pfe_returns = ln(pfe_close/pfe_close[_n-1])
    
	* Sanctions
	  gen tsanction3 = 1 if date > td("04jun1989") & date < td("25may1990")
	  recode tsanction3(.=0)
	
    * Bond Market Crash
	  gen f13minicrash = 1 if date > td("11oct1989") & date < td("17oct1989")
	  recode f13minicrash(.=0)
	  
    * Limit Time Period
      keep if date > td("30dec1988") & date < td("01jun1990")
		
    * Set for analysis
	  tsset t
	  
	* Table A.2 Models
	  eststo clear
	  
	  eststo: arch pfe_returns, ar(1)
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch pfe_returns, ar(1) arch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch pfe_returns, ar(1) arch(1,3)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch pfe_returns, ar(1) arch(1) garch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch pfe_returns, ar(1) arch(1,3) het(tsanction3 f13minicrash)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Pfizer 1989) nodep
	  esttab using rawtables/pfizer89.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Pfizer 1989) nodep replace
	
  * Bed Bath and Beyond (2011)
  
    * Clear
	  clear
	  
	* Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen bbby_returns = ln(bbby_close/bbby_close[_n-1])
    
    * Currency Sanctions
      gen csanction = 1 if date > td("11jul2011") & date < td("18oct2011")
      recode csanction(.=0)

    * Debt Debate
      gen debt = 1 if date > td("13may2011") & date < td("01aug2011")
      recode debt(.=0)

    * Black Mondayd		
      gen bmonday = 1 if date > td("01aug2011") & date < td("15aug2011")
      recode bmonday(.=0)

    * Limit Time Period
      keep if date > td("31dec2010") & date < td("03jan2012")	
		
    * Set for analysis
	  tsset t
	  
	* Table A.3 Models
	  eststo clear
	  
	  eststo: arch bbby_returns
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch bbby_returns, ar(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch bbby_returns, ar(1) ma(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch bbby_returns, ar(1) het(bmonday debt csanction)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch bbby_returns, het(bmonday debt csanction)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Bed Bath and Beyond 2011) nodep
	  esttab using rawtables/bbby2011.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Bed Bath and Beyond 2011) nodep replace
  
  * Gap (2011)
  
    * Clear
	  clear
	  
	* Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen gps_returns = ln(gps_close/gps_close[_n-1])
    
    * Currency Sanctions
      gen csanction = 1 if date > td("11jul2011") & date < td("18oct2011")
      recode csanction(.=0)

    * Debt Debate
      gen debt = 1 if date > td("13may2011") & date < td("01aug2011")
      recode debt(.=0)

    * Black Mondayd		
      gen bmonday = 1 if date > td("01aug2011") & date < td("15aug2011")
      recode bmonday(.=0)
	  
	* Gap Report
      gen report = 1 if date > td("18may2011") & date < td("24may2011")
      recode report(.=0)  

    * Limit Time Period
      keep if date > td("31dec2010") & date < td("03jan2012")	
		
    * Set for analysis
	  tsset t
	  
	* Table A.4 Models
	  eststo clear
	  
	  eststo: arch gps_returns
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch gps_returns, ar(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch gps_returns, ar(1) ma(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch gps_returns, ar(1) ma(1) het(bmonday debt csanction report)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch gps_returns, ar(1) het(bmonday debt csanction report)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch gps_returns, het(bmonday debt csanction report)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Gap 2011) nodep
	  esttab using rawtables/gps2011.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday report GapReport) nomtitles title(Gap 2011) nodep replace
	  
  * Apple (1989)
  
    * Clear
	  clear
	  
	* Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen aapl_returns = ln(aapl_close/aapl_close[_n-1])
    
	* Sanctions
	  gen tsanction3 = 1 if date > td("04jun1989") & date < td("25may1990")
      recode tsanction3(.=0)

	* Bond Market Crash    
	  gen f13minicrash = 1 if date > td("11oct1989") & date < td("17oct1989")
	  recode f13minicrash(.=0)
	  
	* Limit Time Period
      keep if date > td("30dec1988") & date < td("01jun1990")
		
    * Set for analysis
	  tsset t
	  
	* Table A.5 Models  
      eststo clear	
	
	  eststo: arch aapl_returns, ar(1)
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch aapl_returns, ar(1) ma(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch aapl_returns, ar(1) arch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch aapl_returns, ar(1) arch(1) garch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch aapl_returns, ar(1) het(tsanction3 f13minicrash)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch aapl_returns, het(tsanction3 f13minicrash)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Apple 1989) nodep
	  esttab using rawtables/apple1989.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Apple 1989) nodep replace
	  
  * Apple (2011)
  
    * Clear
	  clear
	  
	* Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen aapl_returns = ln(aapl_close/aapl_close[_n-1])
    
    * Currency Sanctions
      gen csanction = 1 if date > td("11jul2011") & date < td("18oct2011")
      recode csanction(.=0)

    * Debt Debate
      gen debt = 1 if date > td("13may2011") & date < td("01aug2011")
      recode debt(.=0)

    * Black Mondayd		
      gen bmonday = 1 if date > td("01aug2011") & date < td("15aug2011")
      recode bmonday(.=0)

    * Limit Time Period
      keep if date > td("31dec2010") & date < td("03jan2012")	
		
    * Set for analysis
	  tsset t
	  
	* Table A.6 Models
	  eststo clear
	  
	  eststo: arch aapl_returns, ar(1)
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch aapl_returns, ar(1) ma(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch aapl_returns, ar(1) ma(1) arch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch aapl_returns, ar(1) ma(1) arch(1) garch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch aapl_returns, ar(1) ma(1) arch(1) garch(1) het(bmonday debt csanction)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
      esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Apple 2011) nodep
      esttab using rawtables/apple2011.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Apple 2011) nodep	replace  
  
  * Dell (1989)	
  
    * Clear
	  clear
	  
	* Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen dell_returns = ln(dell_close/dell_close[_n-1])
    
	* Sanctions
	  gen tsanction3 = 1 if date > td("04jun1989") & date < td("25may1990")
      recode tsanction3(.=0)

	* Bond Market Crash    
	  gen f13minicrash = 1 if date > td("11oct1989") & date < td("17oct1989")
	  recode f13minicrash(.=0)
	  
	* Limit Time Period
      keep if date > td("30dec1988") & date < td("01jun1990")
		
    * Set for analysis
	  tsset t
	  
	* Table A.7 Models  
      eststo clear	
	  
	  eststo: arch dell_returns, ar(1)
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dell_returns, ar(1) ma(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dell_returns, ar(1) ma(1) arch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dell_returns, ar(1) arch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dell_returns, arch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dell_returns, arch(1) garch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dell_returns, arch(1) het(tsanction3 f13minicrash)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Dell 1989) nodep
	  esttab using rawtables/dell1989.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Dell 1989) nodep replace
	  
  * Dell (2011)
  
    * Clear
	  clear
	  
	* Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen dell_returns = ln(dell_close/dell_close[_n-1])
    
    * Currency Sanctions
      gen csanction = 1 if date > td("11jul2011") & date < td("18oct2011")
      recode csanction(.=0)

    * Debt Debate
      gen debt = 1 if date > td("13may2011") & date < td("01aug2011")
      recode debt(.=0)

    * Black Mondayd		
      gen bmonday = 1 if date > td("01aug2011") & date < td("15aug2011")
      recode bmonday(.=0)

    * Limit Time Period
      keep if date > td("31dec2010") & date < td("03jan2012")	
		
    * Set for analysis
	  tsset t
	  
	* Table A.8 Models
	  eststo clear
	  
	  eststo: arch dell_returns, ar(1)
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dell_returns, ar(1) ma(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dell_returns, ar(1) ma(1) arch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dell_returns, ar(1) arch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dell_returns, arch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dell_returns, arch(1) garch(1)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dell_returns, arch(1) garch(1) het(bmonday debt csanction)
	  
	  drop e v s se se2
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Dell 2011) nodep
	  esttab using rawtables/dell2011.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Dell 2011) nodep replace
	  
  * Dow (1989)
  
    * Clear
	  clear
	  
	* Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen dd_returns = ln(dd_close/dd_close[_n-1])
    
	* Sanctions
	  gen tsanction3 = 1 if date > td("04jun1989") & date < td("25may1990")
      recode tsanction3(.=0)

	* Bond Market Crash    
	  gen f13minicrash = 1 if date > td("11oct1989") & date < td("17oct1989")
	  recode f13minicrash(.=0)
	  
	* Limit Time Period
      keep if date > td("30dec1988") & date < td("01jun1990")
		
    * Set for analysis
	  tsset t
	  
	* Table A.9 Models  
      eststo clear	
	  
	  eststo: arch dd_returns, ar(1)
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dd_returns, ar(1) ma(1) 
	  
	  drop e v s se se2

	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dd_returns, ar(1) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dd_returns, ar(1) arch(1) garch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dd_returns, ar(1) arch(1,3)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dd_returns, ar(1) arch(1,3) het(tsanction3 f13minicrash)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Dow 1989) nodep
	  esttab using rawtables/dow1989.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Dow 1989) nodep replace
	  
  * Dow (2011)
  
    * Clear
	  clear
	  
	* Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen dd_returns = ln(dd_close/dd_close[_n-1])
	  
    * Currency Sanctions
      gen csanction = 1 if date > td("11jul2011") & date < td("18oct2011")
      recode csanction(.=0)

    * Debt Debate
      gen debt = 1 if date > td("13may2011") & date < td("01aug2011")
      recode debt(.=0)

    * Black Mondayd		
      gen bmonday = 1 if date > td("01aug2011") & date < td("15aug2011")
      recode bmonday(.=0)

    * Limit Time Period
      keep if date > td("31dec2010") & date < td("03jan2012")	
		
    * Set for analysis
	  tsset t
	  
	* Table A.10 Models
	  eststo clear
	  
	  eststo: arch dd_returns, ar(1)
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dd_returns, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dd_returns, ar(1,2) ma(1,2)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dd_returns, ar(1,2) ma(1,2) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dd_returns, ar(1,2) ma(1,2) arch(1) garch(1) 
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dd_returns, ar(1,2) ma(1,2) arch(1,2) garch(1) 
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dd_returns, ar(1,2) ma(1,2) arch(1,2) garch(1) het(bmonday debt csanction)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dd_returns, ar(9, 25) arch(1,2) garch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch dd_returns, ar(9, 25) arch(1,2) garch(1) het(bmonday debt csanction)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Dow Chemical 2011) nodep  
	  esttab using rawtables/dow2011.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Dow Chemical 2011) nodep replace 
	
  * PPG (1989)
  
    * Clear
	  clear
	  
	* Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen ppg_returns = ln(ppg_close/ppg_close[_n-1])
    
	* Sanctions
	  gen tsanction3 = 1 if date > td("04jun1989") & date < td("25may1990")
      recode tsanction3(.=0)

	* Bond Market Crash    
	  gen f13minicrash = 1 if date > td("11oct1989") & date < td("17oct1989")
	  recode f13minicrash(.=0)
	  
	* Limit Time Period
      keep if date > td("30dec1988") & date < td("01jun1990")
		
    * Set for analysis
	  tsset t
	  
	* Table A.11 Models  
      eststo clear	
	  
	  eststo: arch ppg_returns, ar(1) ma(1)
    
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch ppg_returns, ar(11,16)
	
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch ppg_returns, ar(11,16) arch(1) garch(1)
	
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch ppg_returns, ar(11,16) arch(1,2) garch(1,2)
	
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch ppg_returns, ar(11,16) arch(1,2) garch(1)
	
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch ppg_returns, ar(11,16) arch(1) garch(1)
	
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch ppg_returns, ar(11,16) arch(1) garch(1) het(tsanction3 f13minicrash)
	
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch ppg_returns, ar(11,16) het(tsanction3 f13minicrash)
	
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch ppg_returns, ar(1,11,13,16) het(tsanction3 f13minicrash)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(PPG 1989) nodep
	  esttab using rawtables/ppg1989.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(PPG 1989) nodep	replace
	  
  * PPG (2011)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen ppg_returns = ln(ppg_close/ppg_close[_n-1])
	  
    * Currency Sanctions
      gen csanction = 1 if date > td("11jul2011") & date < td("18oct2011")
      recode csanction(.=0)

    * Debt Debate
      gen debt = 1 if date > td("13may2011") & date < td("01aug2011")
      recode debt(.=0)

    * Black Mondayd		
      gen bmonday = 1 if date > td("01aug2011") & date < td("15aug2011")
      recode bmonday(.=0)

    * Limit Time Period
      keep if date > td("31dec2010") & date < td("03jan2012")	
		
    * Set for analysis
	  tsset t
	  
	* Table A.12 Models
	  eststo clear	  
	  
	  eststo: arch ppg_returns, ar(1)
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch ppg_returns, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch ppg_returns, ar(1) arch(2)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch ppg_returns, arch(2)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch ppg_returns, arch(2,7)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch ppg_returns, arch(2,7) het(bmonday debt csanction)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(PPG 2011) nodep  
	  esttab using rawtables/ppg2011.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(PPG 2011) nodep replace
	  
  * Walmart (1989)
  
    * Clear
	  clear
	  
	* Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen wmt_returns = ln(wmt_close/wmt_close[_n-1])
    
	* Sanctions
	  gen tsanction3 = 1 if date > td("04jun1989") & date < td("25may1990")
      recode tsanction3(.=0)

	* Bond Market Crash    
	  gen f13minicrash = 1 if date > td("11oct1989") & date < td("17oct1989")
	  recode f13minicrash(.=0)
	  
	* Limit Time Period
      keep if date > td("30dec1988") & date < td("01jun1990")
		
    * Set for analysis
	  tsset t
	  
	* Table A.13 Models  
      eststo clear	
  
	  eststo: arch wmt_returns, ar(1)
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch wmt_returns, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch wmt_returns, ar(1) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch wmt_returns, ar(1) ma(1) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch wmt_returns, ar(1) arch(1) het(tsanction3 f13minicrash)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch wmt_returns, arch(1) het(tsanction3 f13minicrash)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch wmt_returns, ar(17) arch(1) het(tsanction3 f13minicrash)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Walmart 1989) nodep
      esttab using rawtables/walmart1989.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Walmart 1989) nodep replace

  * Walmart (2011)
	
	* Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen wmt_returns = ln(wmt_close/wmt_close[_n-1])
	  
    * Currency Sanctions
      gen csanction = 1 if date > td("11jul2011") & date < td("18oct2011")
      recode csanction(.=0)

    * Debt Debate
      gen debt = 1 if date > td("13may2011") & date < td("01aug2011")
      recode debt(.=0)

    * Black Mondayd		
      gen bmonday = 1 if date > td("01aug2011") & date < td("15aug2011")
      recode bmonday(.=0)

    * Limit Time Period
      keep if date > td("31dec2010") & date < td("03jan2012")	
		
    * Set for analysis
	  tsset t
	  
	* Table A.14 Models  
      eststo clear	
	  
	  eststo: arch wmt_returns, ar(1)
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch wmt_returns, ar(1) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch wmt_returns, arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch wmt_returns, arch(1) het(bmonday debt csanction)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Walmart 2011) nodep  
	  esttab using rawtables/walmart2011.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Walmart 2011) nodep replace
	  
  * Duke (1989)
  
    * Clear
	  clear
	  
	* Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen duke_returns = ln(duke_close/duke_close[_n-1])
    
	* Sanctions
	  gen tsanction3 = 1 if date > td("04jun1989") & date < td("25may1990")
      recode tsanction3(.=0)

	* Bond Market Crash    
	  gen f13minicrash = 1 if date > td("11oct1989") & date < td("17oct1989")
	  recode f13minicrash(.=0)
	  
	* Limit Time Period
      keep if date > td("30dec1988") & date < td("01jun1990")
		
    * Set for analysis
	  tsset t
	  
	* Table A.15 Models  
      eststo clear	
  
	  eststo: arch duke_returns, ar(1)
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch duke_returns, ar(9)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch duke_returns, ar(9) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch duke_returns, ar(9) arch(1,8)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch duke_returns, ar(9) arch(1,8) het(tsanction3 f13minicrash)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Duke 1989) nodep
	  esttab using rawtables/duke1989.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Duke 1989) nodep replace
	  
  * Duke (2011)	 
  
	* Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen duke_returns = ln(duke_close/duke_close[_n-1])
	  
    * Currency Sanctions
      gen csanction = 1 if date > td("11jul2011") & date < td("18oct2011")
      recode csanction(.=0)

    * Debt Debate
      gen debt = 1 if date > td("13may2011") & date < td("01aug2011")
      recode debt(.=0)

    * Black Mondayd		
      gen bmonday = 1 if date > td("01aug2011") & date < td("15aug2011")
      recode bmonday(.=0)

    * Limit Time Period
      keep if date > td("31dec2010") & date < td("03jan2012")	
		
    * Set for analysis
	  tsset t
	  
	* Table A.16 Models  
      eststo clear	
	  
	  eststo: arch duke_returns, ar(1) 
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch duke_returns, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch duke_returns, ma(1) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch duke_returns, ar(1) arch(9) het(bmonday debt csanction)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch duke_returns, ma(1) arch(9) het(bmonday debt csanction)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Duke 2011) nodep  
	  esttab using rawtables/duke2011.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Duke 2011) nodep replace
	  
  * Deere (1989)
  
    * Clear
	  clear
	  
	* Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen de_returns = ln(de_close/de_close[_n-1])
    
	* Sanctions
	  gen tsanction3 = 1 if date > td("04jun1989") & date < td("25may1990")
      recode tsanction3(.=0)

	* Bond Market Crash    
	  gen f13minicrash = 1 if date > td("11oct1989") & date < td("17oct1989")
	  recode f13minicrash(.=0)
	  
	* Limit Time Period
      keep if date > td("30dec1988") & date < td("01jun1990")
		
    * Set for analysis
	  tsset t
	  
	* Table A.17 Models  
      eststo clear	
	  
	  eststo: arch de_returns, ar(1)
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch de_returns, ar(3)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch de_returns, ar(3) ma(5)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch de_returns, ar(3,5)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch de_returns, ar(3,5,8)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch de_returns, ar(3,5,8) het(tsanction3 f13minicrash)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Deere 1989) nodep
	  esttab using rawtables/deere1989.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant tsanction3 Sanction f13minicrash MiniCrash) nomtitles title(Deere 1989) nodep replace
	  
  * Deere (2011)	  
  
	* Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen de_returns = ln(de_close/de_close[_n-1])
	  
    * Currency Sanctions
      gen csanction = 1 if date > td("11jul2011") & date < td("18oct2011")
      recode csanction(.=0)

    * Debt Debate
      gen debt = 1 if date > td("13may2011") & date < td("01aug2011")
      recode debt(.=0)

    * Black Mondayd		
      gen bmonday = 1 if date > td("01aug2011") & date < td("15aug2011")
      recode bmonday(.=0)

    * Limit Time Period
      keep if date > td("31dec2010") & date < td("03jan2012")	
		
    * Set for analysis
	  tsset t
	  
	* Table A.18 Models  
      eststo clear	
	  
	  eststo: arch de_returns, ar(1)
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch de_returns, ar(3,5)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch de_returns, ar(3,5) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch de_returns, ar(3,5) arch(1,2)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch de_returns, ar(3,5) arch(1) garch(1) 
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch de_returns, ar(3,5) arch(2) garch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch de_returns, ar(3,5) arch(1,2,13) 
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch de_returns, ar(3,5) arch(2) garch(1) het(bmonday debt csanction)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Deere 2011) nodep  
	  esttab using rawtables/deere2011.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant csanction Sanction debt DebtDebate bmonday BlackMonday) nomtitles title(Deere 2011) nodep replace
	  
  * Exxon Mobile 1992 (Columbia)
  
  	* Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen xom_returns = ln(xom_close/xom_close[_n-1])
	  
	* Sanctions Variable
	  gen sanct = 1 if date > td("16jan1992") & date < td("30jan1992")
	  recode sanct(.=0)
  
	  gen sanc = 1 if date > td("29jan1992")
	  recode sanc(.=0)  
	  
    * Limit Time Period
	  keep if date > td("01jun1991") & date < td("01jun1992")
	  
    * Set for analysis
	  tsset t
	  
	* Table A.19 Models  
	  eststo clear	
	  
	  eststo: arch xom_returns
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch xom_returns, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch xom_returns, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch xom_returns, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch xom_returns, ar(1) ma(1) het(sanc sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions sanct SanctionThreats) nomtitles title(Exxon Mobil 1992) nodep  
	  esttab using rawtables/exxon1992.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions sanct SanctionThreats) nomtitles title(Exxon Mobil 1992) nodep replace
  
  * Halliburton 1992 (Columbia)
  
   	* Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen hal_returns = ln(hal_close/hal_close[_n-1])
	  
	* Sanctions Variable
	  gen sanct = 1 if date > td("16jan1992") & date < td("30jan1992")
	  recode sanct(.=0)
  
	  gen sanc = 1 if date > td("29jan1992")
	  recode sanc(.=0)  
	  
    * Limit Time Period
	  keep if date > td("01jun1991") & date < td("01jun1992")
	  
    * Set for analysis
	  tsset t
	  
	* Table A.20 Models  
	  eststo clear
	  
	  eststo: arch hal_returns
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch hal_returns, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch hal_returns, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch hal_returns, ar(1) ma(1) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch hal_returns, ar(1) ma(1) arch(1) garch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch hal_returns, ar(1) ma(1) arch(1) garch(1) het(sanc sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch hal_returns, het(sanc sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions sanct SanctionThreats) nomtitles title(Halliburton 1992) nodep  
	  esttab using rawtables/halliburton1992.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions sanct SanctionThreats) nomtitles title(Halliburton 1992) nodep replace
  
  * Colgate-Palmolive 1993 (Romania)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_cl = ln(cl_close/cl_close[_n-1])	 
	  
	* Sanctions Variable	
	  gen sanc = 1 if date > td("26jan1993")
	  recode sanc(.=0)	 
	  
	* Limit Time Period
	  keep if date > td("31jan1991") & date < td("01jan1994")
	  
	* Set for analysis
	  tsset t
	  
	* Table A.21 Models
	  eststo clear
	  
	  eststo: arch returns_cl
	  	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_cl, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_cl, ar(5)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_cl, ma(5)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_cl, ar(5) het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Colgate-Palmolive 1993 (Romania)) nodep  
	  esttab using rawtables/colgate1993.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Colgate-Palmolive 1993 (Romania)) nodep replace

  * Kimberly-Clark 1993 (Romania)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_kmb = ln(kmb_close/kmb_close[_n-1])	 
	  
	* Sanctions Variable	
	  gen sanc = 1 if date > td("26jan1993")
	  recode sanc(.=0)	 
	  
	* Limit Time Period
	  keep if date > td("31jan1991") & date < td("01jan1994")
	  
	* Set for analysis
	  tsset t
	  
	* Table A.22 Models	  
	  eststo clear
	  
	  eststo: arch returns_kmb
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_kmb, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_kmb, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_kmb, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_kmb, arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_kmb, ar(1) ma(1) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_kmb, arch(1) het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Kimberly-Clark 1993 (Romania)) nodep  
	  esttab using rawtables/kimberly1993.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Kimberly-Clark 1993 (Romania)) nodep replace
  
  * Disney 1997 (France)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_dis = ln(dis_close/dis_close[_n-1])	 
	  
	* Sanctions Variable	
	  gen sanct = 1 if date > td("28sep1997") & date < td("18may1998")
	  recode sanct(.=0)
	  
	* Limit Time Period
	  keep if date > td("31dec1996") & date < td("01jan1998")
	  
	* Set for analysis
	  tsset t
	  
	* Table A.23 Models
	  eststo clear
	  
	  eststo: arch returns_dis
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_dis, ar(2)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_dis, ar(2) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_dis, ar(2) ma(1) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_dis, ar(2) ma(1) arch(1) garch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_dis, ar(2) ma(1) arch(1) het(sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_dis, ar(2) ma(1) het(sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_dis, ar(2) het(sanct)
	
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Disney 1997 (France)) nodep  
	  esttab using rawtables/disney1997.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Disney 1997 (France)) nodep replace

  * AT&T Corp (France)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_t = ln(t_close/t_close[_n-1])	 
	  
	* Sanctions Variable	
	  gen sanct = 1 if date > td("28sep1997") & date < td("18may1998")
	  recode sanct(.=0)
	  
	* Limit Time Period
	  keep if date > td("31dec1996") & date < td("01jan1998")
	  
	* Set for analysis
	  tsset t
	  
	* Table A.24 Models
	  eststo clear
	  
	  eststo: arch returns_t
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_t, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_t, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_t, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_t, ar(3)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_t, ar(3) arch(1) 
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_t, ar(3) arch(1) garch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_t, ar(3) het(sanct)

	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanct SanctionsThreat) nomtitles title(ATT Corp 1997 (France)) nodep  
	  esttab using rawtables/atandt1997.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanct SanctionsThreat) nomtitles title(ATT Corp 1997 (France)) nodep replace
	  
  * Ford 1991 (Germany)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_f = ln(f_close/f_close[_n-1])	 
	  
	* Sanctions Variable	
	  gen sanct = 1 if date > td("31jan1991") & date < td("03mar1991")
	  recode sanct(.=0)  
	  
	* Limit Time Period
	  keep if date > td("31dec1990") & date < td("01jan1992")
	  
	* Set for analysis
	  tsset t
	  
	* Table A.25 Models
	  eststo clear
	  
	  eststo: arch returns_f
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_f, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_f, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_f, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_f, het(sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanct SanctionsThreat) nomtitles title(Ford 1991 (Germany)) nodep  
	  esttab using rawtables/ford1991.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanct SanctionsThreat) nomtitles title(Ford 1991 (Germany)) nodep replace
	 
  * TJX Companies 1991 (Germany)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_tjx = ln(tjx_close/tjx_close[_n-1])

	* Sanctions Variable	
	  gen sanct = 1 if date > td("31jan1991") & date < td("03mar1991")
	  recode sanct(.=0)  
	  
	* Limit Time Period
	  keep if date > td("31dec1990") & date < td("01jan1992")
	  
	* Set for analysis
	  tsset t
	  
	* Table A.26 Models
	  eststo clear
	  
	  eststo: arch returns_tjx
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_tjx, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_tjx, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_tjx, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_tjx, ma(3)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_tjx, ma(3) het(sanct)
	  	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanct SanctionsThreat) nomtitles title(TJX Companies 1991 (Germany)) nodep  
	  esttab using rawtables/tjx1991.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanct SanctionsThreat) nomtitles title(TJX Companies 1991 (Germany)) nodep replace
	  	 
  * Advanced Micro Devices, Inc. 1997 (Japan)

    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_amd  = ln(amd_close/amd_close[_n-1])	  

	* Sanctions Variable	
	  gen sanc = 1 if date > td("25sep1997")
	  recode sanc(.=0)    
	  
	* Limit Time Period
	  keep if date > td("31dec1996") & date < td("01jan1998")	  
	  
	* Set for analysis
	  tsset t
	  
	* Table A.27 Models
	  eststo clear
	  
	  eststo: arch returns_amd
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_amd, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_amd, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_amd, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_amd, ar(1,4)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_amd, ar(1,4) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_amd, ar(1,4) arch(1) garch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_amd, ar(1,4) arch(1, 17)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_amd, ar(1,4) arch(1, 17) het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Advanced Micro Devices, Inc. 1997 (Japan)) nodep  
	  esttab using rawtables/amd1997.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Advanced Micro Devices, Inc. 1997 (Japan)) nodep replace

  * Analog Devices, Inc. 1997 (Japan)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_adi  = ln(adi_close/adi_close[_n-1])	  

	* Sanctions Variable	
	  gen sanc = 1 if date > td("25sep1997")
	  recode sanc(.=0)    
	  
	* Limit Time Period
	  keep if date > td("31dec1996") & date < td("01jan1998")	  
	  
	* Set for analysis
	  tsset t
	  
	* Table A.28 Models
	  eststo clear
	  
	  eststo: arch returns_adi
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_adi, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_adi, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_adi, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_adi, ar(1) ma(1) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_adi, ar(1) ma(1) arch(1) het(sanc)
  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Analog Devices, Inc. 1997 (Japan)) nodep  
	  esttab using rawtables/adi1997.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Analog Devices, Inc. 1997 (Japan)) nodep replace
  
  * Intel Corporation 1997 (Japan)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_intc = ln(intc_close/intc_close[_n-1])		  

	* Sanctions Variable	
	  gen sanc = 1 if date > td("25sep1997")
	  recode sanc(.=0)    
	  
	* Limit Time Period
	  keep if date > td("31dec1996") & date < td("01jan1998")	  
	  
	* Set for analysis
	  tsset t
	  
	* Table A.29 Models
	  eststo clear  
	  
	  eststo: arch returns_intc
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_intc, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_intc, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_intc, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_intc, het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Intel Corporation 1997 (Japan)) nodep  
	  esttab using rawtables/intel1997.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Intel Corporation 1997 (Japan)) nodep replace 

  * Maxim Integrated Products Inc. 1997 (Japan)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_mxim = ln(mxim_close/mxim_close[_n-1])		  

	* Sanctions Variable	
	  gen sanc = 1 if date > td("25sep1997")
	  recode sanc(.=0)    
	  
	* Limit Time Period
	  keep if date > td("31dec1996") & date < td("01jan1998")	  
	  
	* Set for analysis
	  tsset t
	  
	* Table A.30 Models
	  eststo clear  
	  
	  eststo: arch returns_mxim
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mxim, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mxim, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mxim, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mxim, ar(2)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mxim, ar(2) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mxim, ar(2) arch(1,17)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mxim, ar(2) arch(1,17) het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Maxim Integrated Products Inc. 1997 (Japan)) nodep  
	  esttab using rawtables/mxim1997.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Maxim Integrated Products Inc. 1997 (Japan)) nodep replace 	  

  * Microchip Technology Inc. 1997 (Japan)	  
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_mchp = ln(mchp_close/mchp_close[_n-1])	  

	* Sanctions Variable	
	  gen sanc = 1 if date > td("25sep1997")
	  recode sanc(.=0)    
	  
	* Limit Time Period
	  keep if date > td("31dec1996") & date < td("01jan1998")	  
	  
	* Set for analysis
	  tsset t
	  
	* Table A.31 Models
	  eststo clear  
	  
	  eststo: arch returns_mchp
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mchp, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mchp, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mchp, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mchp, arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mchp, arch(14)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mchp, het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mchp, arch(14) het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Microchip Technology Inc. 1997 (Japan)) nodep  
	  esttab using rawtables/mchp1997.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Microchip Technology Inc. 1997 (Japan)) nodep replace 	  
	  
  * Micron Technology, Inc. 1997 (Japan)	  
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_mu   = ln(mu_close/mu_close[_n-1])	  	  

	* Sanctions Variable	
	  gen sanc = 1 if date > td("25sep1997")
	  recode sanc(.=0)    
	  
	* Limit Time Period
	  keep if date > td("31dec1996") & date < td("01jan1998")	  
	  
	* Set for analysis
	  tsset t
	  
	* Table A.32 Models
	  eststo clear  
	  
	  eststo: arch returns_mu
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mu, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mu, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mu, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mu, arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mu, arch(1) het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mu, arch(1) ar(14,17) het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Micron Technology, Inc. 1997 (Japan)) nodep  
	  esttab using rawtables/mu1997.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Micron Technology, Inc. 1997 (Japan)) nodep replace 	  

  * Qualcomm 1997 (Japan)	
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_qcom = ln(qcom_close/qcom_close[_n-1])	    	  

	* Sanctions Variable	
	  gen sanc = 1 if date > td("25sep1997")
	  recode sanc(.=0)    
	  
	* Limit Time Period
	  keep if date > td("31dec1996") & date < td("01jan1998")	  
	  
	* Set for analysis
	  tsset t
	  
	* Table A.33 Models
	  eststo clear 
	  
	  eststo: arch returns_qcom
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_qcom, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_qcom, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_qcom, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_qcom, het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_qcom, arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_qcom, arch(1) garch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_qcom, arch(1) het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Qualcomm 1997 (Japan)) nodep  
	  esttab using rawtables/qcom1997.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Qualcomm 1997 (Japan)) nodep replace
	  
  * Texas Instruments 1997 (Japan)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_txn  = ln(txn_close/txn_close[_n-1])	  	    	  

	* Sanctions Variable	
	  gen sanc = 1 if date > td("25sep1997")
	  recode sanc(.=0)    
	  
	* Limit Time Period
	  keep if date > td("31dec1996") & date < td("01jan1998")	  
	  
	* Set for analysis
	  tsset t
	  
	* Table A.34 Models
	  eststo clear 
	  
	  eststo: arch returns_txn
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_txn, ar(10)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_txn, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_txn, ar(10) het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_txn, ar(7,10) het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_txn, ar(7,10) arch(1) het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_txn, ar(7,10) arch(1) garch(1) het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_txn, ar(7,10) arch(1,2) garch(1,2) het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Texas Instruments 1997 (Japan)) nodep  
	  esttab using rawtables/txn1997.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Texas Instruments 1997 (Japan)) nodep replace  

  * Xilinix 1997 (Japan)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_xlnx = ln(xlnx_close/xlnx_close[_n-1])	  	    	  

	* Sanctions Variable	
	  gen sanc = 1 if date > td("25sep1997")
	  recode sanc(.=0)    
	  
	* Limit Time Period
	  keep if date > td("31dec1996") & date < td("01jan1998")	  
	  
	* Set for analysis
	  tsset t
	  
	* Table A.35 Models
	  eststo clear 
	  
	  eststo: arch returns_xlnx
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_xlnx, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_xlnx, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_xlnx, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_xlnx, ar(1) ma(1) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_xlnx, ar(1) ma(1) arch(1) het(sanc)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Xilinix 1997 (Japan)) nodep  
	  esttab using rawtables/xlnx1997.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Xilinix 1997 (Japan)) nodep replace 

  * Allergan 1999 (Spain)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_agn = ln(agn_close/agn_close[_n-1])	  	    	  

	* Sanctions Variable	
	  gen sanct = 1 if date > td("05apr1999") & date < td("11may1999")
	  recode sanct(.=0)
	  
	* Limit Time Period
	  keep if date > td("31dec1998") & date < td("01jan2000")	  
	  
	* Set for analysis
	  tsset t
	  
	* Table A.36 Models
	  eststo clear 
	  
	  eststo: arch returns_agn
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_agn, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_agn, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_agn, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_agn, het(sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_agn, ma(1) het(sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Allergan 1999 (Spain)) nodep  
	  esttab using rawtables/agn1999.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Allergan 1999 (Spain)) nodep replace 
    
  * Johnson & Johnson 1999 (Spain)
      
	* Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_jnj = ln(jnj_close/jnj_close[_n-1])	  	    	  

	* Sanctions Variable	
	  gen sanct = 1 if date > td("05apr1999") & date < td("11may1999")
	  recode sanct(.=0)
	  
	* Limit Time Period
	  keep if date > td("31dec1998") & date < td("01jan2000")	  
	  
	* Set for analysis
	  tsset t
	  
	* Table A.37 Models
	  eststo clear 
	  
	  eststo: arch returns_jnj
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_jnj, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_jnj, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_jnj, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_jnj, het(sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_jnj, ma(1) het(sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Johnson & Johnson 1999 (Spain)) nodep  
	  esttab using rawtables/jnj1999.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Johnson & Johnson 1999 (Spain)) nodep replace 

  * Eli Lilly 1999 (Spain)
  
  	* Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_lly = ln(lly_close/lly_close[_n-1]) 	    	  

	* Sanctions Variable	
	  gen sanct = 1 if date > td("05apr1999") & date < td("11may1999")
	  recode sanct(.=0)
	  
	* Limit Time Period
	  keep if date > td("31dec1998") & date < td("01jan2000")	  
	  
	* Set for analysis
	  tsset t
	  
	* Table A.38 Models
	  eststo clear 
	  
	  eststo: arch returns_lly
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_lly, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_lly, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_lly, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_lly, het(sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Eli Lilly 1999 (Spain)) nodep  
	  esttab using rawtables/lly1999.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Eli Lilly 1999 (Spain)) nodep replace 

  * Merck & Co. 1999 (Spain)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_mrk = ln(mrk_close/mrk_close[_n-1]) 	    	  

	* Sanctions Variable	
	  gen sanct = 1 if date > td("05apr1999") & date < td("11may1999")
	  recode sanct(.=0)
	  
	* Limit Time Period
	  keep if date > td("31dec1998") & date < td("01jan2000")	  
	  
	* Set for analysis
	  tsset t
	  
	* Table A.39 Models
	  eststo clear 	  
	  
	  eststo: arch returns_mrk
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mrk, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mrk, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mrk, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mrk, arch(3)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mrk, het(sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_mrk, arch(1) het(sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Merck & Co. 1999 (Spain)) nodep  
	  esttab using rawtables/mrk1999.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Merck & Co. 1999 (Spain)) nodep replace 
	  
  * Perrigo 1999 (Spain)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_prgo = ln(prgo_close/prgo_close[_n-1])	  

	* Sanctions Variable	
	  gen sanct = 1 if date > td("05apr1999") & date < td("11may1999")
	  recode sanct(.=0)
	  
	* Limit Time Period
	  keep if date > td("31dec1998") & date < td("01jan2000")	  
	  
	* Set for analysis
	  tsset t
  
	* Table A.40 Perrigo 1999 (Spain)
	  eststo clear 
	  
	  eststo: arch returns_prgo
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_prgo, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_prgo, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_prgo, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_prgo, ar(1) ma(1) arch(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_prgo, ar(1) ma(1) arch(1,4)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_prgo, ar(1) ma(1) arch(1,4) het(sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Perrigo 1999 (Spain)) nodep  
	  esttab using rawtables/prgo1999.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Perrigo 1999 (Spain)) nodep replace 

  * Pfizer 1999 (Spain)
  
    * Clear
	  clear
	  
    * Load Data
	  use "rcs.dta"
  
    * Generate Returns
	  gen returns_pfe = ln(pfe_close/pfe_close[_n-1])	  

	* Sanctions Variable	
	  gen sanct = 1 if date > td("05apr1999") & date < td("11may1999")
	  recode sanct(.=0)
	  
	* Limit Time Period
	  keep if date > td("31dec1998") & date < td("01jan2000")	  
	  
	* Set for analysis
	  tsset t
  
	* Table A.41 Pfizer 1999 (Spain)
	  eststo clear
	  
	  eststo: arch returns_pfe
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_pfe, ar(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_pfe, ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_pfe, ar(1) ma(1)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_pfe, ar(1,5)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  eststo: arch returns_pfe, ar(1,5) het(sanct)
	  
	  drop e v s se se2	  
	  
	  predict e, residuals
      predict v, variance
      gen s = sqrt(v)
      gen se = e/s
      gen se2 = se^2
      
	  ac se
	  pac se
	  wntestq se
	  
      ac se2
      pac se2
      wntestq se2
	  
	  esttab, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Pfizer 1999 (Spain)) nodep  
	  esttab using rawtables/pfe1999.tex, b(3) se(3) star(* .10 ** .05 *** .01) scalars(aic bic) nogaps compress coeflabels(_cons Constant sanc Sanctions) nomtitles title(Pfizer 1999 (Spain)) nodep replace 
	  
* Close log file
  log close	  
	  
  
  
	  
	  
	  
	  
